Non-ideal subthreshold swing in aligned carbon nanotube transistors due to variable occupancy discrete charge traps

TL;DR

This paper uses simulations to show that discrete charge traps near carbon nanotubes cause non-ideal subthreshold swing, highlighting the importance of trap density and proximity for device performance.

Contribution

It demonstrates that variable occupancy charge traps near nanotubes significantly impact subthreshold swing, providing insights for improving device fabrication.

Findings

Trap density of ~0.5/nm$^2$ affects swing

Proximity of traps near nanotubes is critical

Few traps per nanotube can cause large swing

Abstract

Carbon nanotube transistors have been experimentally demonstrated to reach performance comparable and even surpassing that of silicon transistors. Further improvement requires addressing non-idealities arising from device fabrication that impact performance and reproducibility. One performance metric that determines energy efficiency is the subthreshold swing which is often observed to be 3-4 times larger than the ideal thermal limit. In this work, we present simulations indicating that a discrete number of variable occupancy hole trapping sites can explain the large subthreshold swing. Our simulations indicate that while three-dimensional trap distributions influence the subthreshold swing, only the traps in close proximity to the nanotubes have a significant impact. The results suggest that a density of trapping sites on the order of 0.5/nm$^2$ near the nanotubes is sufficient to significantly increase the subthreshold swing, requiring the removal or passivation of only a few sites per carbon nanotube.